The prosodic characteristics of mild dysarthric speakers and change of speech intelligibility and acceptability by prosodic manipulation
Dysprosody is a one of the most common dysarthric speech deficits. Prosody plays an important role in the speech intelligibility and acceptability deficit of dysarthria. Speech intelligibility and acceptability is evaluated listener’s perceptual judge. Because acoustic methods are useful for quantification of prosodic variables, it is usually necessary that the unit of prosodic analysis is determined perceptually, analysis of dysprosody is the combined use of perceptual and acoustic methods (Lowit-Leuschel and Docherty, 1999).
Maximizing speech intelligibility and acceptability are common goals of treatments for dysarthria(Duffy, 2013; Yorkston et al., 2010). Clear speech(CS) refers to a speaking style wherein talkers voluntarily adjust their habitual speech(HS) to maximize intelligibility for a communication partner. So, CS is recommended for use as a therapy technique for speakers with dysarthria. Rate reduction and increasing vocal loudness of CS are common treatment techniques. Improvements in Intelligibility of CS associated with a slow rate or increased vocal intensity.
The purpose of this study is to investigate prosodic acoustic characteristics in the speech of adults with mild dysarthria through a comparison with the speech of normal speakers and to determine if the prosodic type of resynthesized cue resulted in a significant perceptual change in speech intelligibility and acceptability.
Twelve speakers with mild dysarthria and twelve healthy controls read a sentences. Sentences were read in HS and CS by two groups. Acoustic measures were obtained using Praat. The habitual Sentences recorded by twelve speakers with dysarthria were made by digitally manipulating at three different rate(habitual, a 20% slower rate and a 20% higher rate) and three different loudness(70dB, 75dB, 80dB). Thirty-five listeners rated the sentences for intelligibility and acceptability by 10cm visual analog scale.
The results revealed that, compared to normal speakers, speakers with mild dysarthria showed a slower speech rate, a higher pause frequency and longer pause duration, and lower intensity. But, there were no statistical differences in intensity and pitch variation in two groups.
Relative to HS, the CS was associated with different acoustic changes for speech rate, intensity and pitch in normal group. But, there were no acoustic changes in dysarthric group except pitch variation.
A faster speech rate than habitual and slower rate resulted in significant improvement in speech intelligibility and acceptability ratings(p<.001). An increased vocal loudness resulted in no significant improvement in speech intelligibility ratings. In speech acceptability ratings, an increased vocal loudness(75, 80dB) resulted in significant improvement(p<.05). Speech rate and loudness changes in the prosodic properties of speech may contribute to improved speech intelligibility and acceptability. Specially in rate, mild dysarthic speakers would benefit from using a faster speech rate than they used in.
The results also showed mild correlations between speech intelligibility and speech acceptability. Speech acceptability does not always correlate with speech intelligibility measures.
As this study used speech samples that were rate and loudness altered by artificial means, investigation into the effects upon speech intelligibility and acceptability are needed using speakers who use self-induced rate changes. Further, if rate and loudness changes are deemed to improve the speech produced by this speakers, it needs to be determined of they can be trained to sustained the most beneficial rate and loudness.